Skip to content
Synthica Journal ISSN pending Submit your research
Analysis Computer Science Open access

Benchmarking Small Language Models for Math Word Problems

Ravi Shah1, Mei Tan1

  1. 1Synthica Research Group

Synthica Journal 1, 1–14 (2025) · Volume 1, Issue 1

Received 2 November 2024 Accepted 18 February 2025 Published 15 March 2025

DOI https://doi.org/10.55555/synthica.2025.0002

5,400 Accesses 6 Citations 27 Altmetric

Download PDF

Abstract

We evaluate sub-3B parameter models on GSM8K-style problems and analyze failure modes.

Keywords

  • language models
  • mathematical reasoning
  • benchmarking

License

Open access — This article is licensed under CC BY 4.0 (Creative Commons Attribution 4.0 International), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, provided the original author(s) and source are credited.